I don’t think the sense-datum theorist can claim to be a direct realist; the arguments for sense-data usually establish that we perceive sense-data. Also, the term ’sense-data’ itself just means ‘what is given to the senses’. Representationalism, however, is consistent with direct realism, as one need not see the representations [it may also be consistent with sense-data views: indirect realist sense-data views used to occasionally be called representational views, because the sense-data themselves represents).

A view which does oppose representationalism is naive realism; that construes experience as a relational state; one being in which requires one be related to the elements of the sense that one perceive. Representationalism denies this; representations do not require the existence of what they represent. Is that the dispute you’re getting at?

That’s why I think that the what direct realist actually negates is the sensibility of phenomenal/conscious/visual or experience taken as something which a)represents the world (veridically or not) and b)which has what it is likeness (qualitative) characteristic to it.

Also, I want to point is that the issue is not directly related with the issue if one is physicalist or not. I’m inclined for example towards direct realism and denying such thing as phenomenal experience, but I am not physicalist. I think that “I see an apple” is describing a situation to which one can give further physical predicates, but that “seeing an apple” predicate of a situation is not reducible to the physical predicates. That is, I agree that when we give predicate “I am seeing an apple” to a situation, we can also give predicate “photons are bouncing off off the surface of the apple, getting focused by the lenses of my eyes, fall on the retina, etc..”, but I don’t think that those predicates are equivalent. So, just wanted to say that one can be direct realist, and not be physicalist.

I kind of think you are on to something with your second point, and I think Gibson, amongst others, probably Heidegger, would agree with you. Perception is merely a brain/body/world event. There is no subjective locus of experience for direct realism. There are merely physical events and I suspect it is our language that constructs the narrative for us to talk about “subjectivity” and “phenomenal experience”.

In my thinking about those things, I’ve concluded two things:

1.There is nothing problematic in saying that ‘we can see the tree’. Whatever is included in the process, like photons bouncing off the tree, the sensory organs registering them, the processing going on in the brain, and EVEN if there is sense-data or something, those would be just parts of the “seeing the tree”. So, I think there is no important distinction there to draw between ‘direct’ and ‘indirect’ realism, and that both can happily agree that “we see the tree (directly)”. (Basically to say that what we see is sense-datum or representation, is to misuse the word ’see’.)

2.So where would be the distinction between the direct realism and indirect one? How to pin down what is direct realism them? I’ve been thinking about it, and I think it is this… Direct realism denies (or should deny) that there is such thing as phenomenal experience. It should say that using ‘experience’ in such sense is nothing but product of philosophers’ theory, and that in proper sense ‘experience’ is used to refer to the events in the world, in which we take part, in which we are aware of those events (or parts of them), and from which we are affected somehow, or from which we learn something.

So, basically I think that the direct realism vs. representationalism issue can’t be framed in the terms of ‘direct’ or ‘indirect’ seeing something, but if those accept or deny that such thing as ‘phenomenal experience’ (or ‘conscious experience’) exist.